3 reasons why your financial institution hasn’t adopted AI

Posted: 17 November 2017 | By Charlie Moloney

It can be frustrating when adoption doesn't pan out as hoped

Financial institutions (FI) can transform their customer experience and remain relevant by investing in artificial intelligence (AI), as Access AI revealed in a feature report on ATB Financial in October.

So, what’s stopping them?

Clara Durodié is the CEO and founder of Cognitive Finance Group, an AI consulting and investment firm, which helps FIs to achieve next level business growth through AI transformation.

She explains the key blockers facing FIs today:

1. lack of understanding of artificial intelligence at the Board level

“In my opinion, a strong mandate for adoption of AI comes from the Board level”, Clara told Access AI, “It’s a mainly top down process.”

“Board members need to have the intellectual curiosity to dive into this information, build their own knowledge, and with that knowledge they can confidently redesign their business strategy. With the same knowledge, they are also empowered to select  an impartial AI advisor to help them further, in plain English, adoption, selection and implementation of AI. Be aware of well-known consulting firms, though. They will always want to sell their own products, so no impartiality.”`

“However, in practical terms it is difficult because boards of FIs have so much responsibility. One Board ‘pack’ [required reading before any Board meeting] is about 400 pages. That is usually distilled from a total of 2,000 – 3,000 pages. They have lots of information to deal with at each Board meeting.”

“AI technology is intimidating because it’s new to the business world, and there is currently no structured way to help educate Boards on this topic. That’s why our firm has developed a dedicated AI for Boards masterclass in addition to our data science delivery. ”

“Also, some see AI as something that is going to become relevant and important in five to ten years’ time. So, this perception can make AI adoption a second priority for them.”

“But a lack of knowledge and understanding at Board level is the main blocker for making decisions.”

2. Too much focus on return on investment (ROI)

Clara told us companies are often focusing on short term ROI, rather than looking at the long term strategic benefits that AI will bring.

“Nobody really knows in 2017 what the ROI for AI technologies will be”, she told us, “even though a lot of people claim that they do.”

“If I saw a company or a consultant put an exact figure for the ROI a piece of AI tech will give, I would not let our team support that figure.”

“Although the ROI for AI is not clear at this time, what is clear is that in the future it will not be possible to compete unless you have adopted this technology. “

“Naturally, we need some sort of metrics to measure the business benefits from adopting AI. ROI in the current fiscal year may not always be the right measure of success. Take the example of Goldman Sachs’s Marcus lending platform. They built it from scratch, and their measure of success was to reach $1bn loans in 12 months. They seem to have reached that milestone in 9 months and moreover doubled it to $2bn in 12 months. They would appear to have exceeded their own measure of success.”

3. The vicious cycle of delay

“The longer the adoption of AI technology is delayed, the harder it will be”, Clara told us.

“Delaying the adoption of the language and the narrative explaining why this technology is being introduced will have negative impacts on re-training staff”.

“What we are seeing is businesses leaving it until late, then they finally start adopting RPA (Robotic Process Automation) which not only is just an quick band-aid but also doesn’t deliver Intelligent Process Automation (automation with AI) which is where the direction is and the resources should be directed to”.

“Then, some start letting people go because they think they don’t have time to retrain them.”

“Part of the advice that we give FIs is to design a joined-up A.I. strategy, which would also incorporate a strategy to retrain as many people as you possibly can so that you can remain relevant”.

“It takes between three to six months to train somebody to use Python. Moreover, it takes time to deploy AI projects so you do need to start early.

So, to distil, where do you start with AI adoption?

“Bring in an impartial AI consultancy which can help you find the best value for money AI solutions globally and help your Board redesign the growth strategy to incorporate AI .”

You can meet Clara and hear her speak live in Spitalfields, London during a free Access AI event being held this Thursday at 8am. RSVP here.

Related industries

Related functions